Model Path Selection

The "kisselmentch2021" folder contains R code to reproduce the simulations, applications, and motivation examples in our upcoming Forward Stability and Model Path Selection paper.

The "ModelPath" folder can be installed as an R package. It provides tools for performing MPS.

R Package Install

devtools::install_github(repo='nkissel/MPS', subdir='ModelPath')

Getting started

For an example of how to use MPS

library(MASS)
n<-1000
p<-10
x<-mvrnorm(n,rep(0,p),diag(p))
signal<-rowSums(x[,1:3])
noise<-rnorm(n,0,1/4)
y<-signal+noise
mydata<-data.frame(x,y)
mps1<-full.select.gen(myframe=mydata,resp.name='y',depth=3,
   r=100,model='lm',condense=FALSE)
mps1
build.tree(mps1) #graph

#merged paths
mps2<-full.select.gen(myframe=mydata,resp.name='y',depth=3,
   r=100,model='lm',condense=TRUE)
mps2
build.tree(mps2) #graph